The specific aim of the proposed research is to investigate the inhibitory activity of substituted canthin-6-one and canthin-5, 6-dione indole alkaloids against human cAMP phosphodiesterase (PDE4) isoenzymes. In the past canthin-6-one indole alkaloids have been shown to elicit inhibitory activity against bovine PDE4 but the particular molecular features necessary for optimum inhibitory activity against human isoforms and the development of a quantitative structure activity relationship to guide the development of new molecular targets has not been explored. The aims of this project are three-fold: 1) synthesize a series of canthin-6-one and canthin-5, 6-done indole alkaloid derivatives, 2) screen these compounds for inhibitory activity against human PDE4 isoforms, and 3) develop a three-dimensional quantitative structure activity relationship so as to provide a direction for further study. Canthinone indole alkaloids are natural products, which have been isolated from various sources. Although canthin-6-one derivatives have been prepared in the past, a direct methodology by which compounds of varied substitution patterns might be prepared has yet to be put forth. We propose to use appropriately substituted tryptamines and the Pictet-Spengler condensation as key steps in the synthesis of these compounds. Synthesis of these compounds will allow pharmacological screening against human recombinant PDE4 isoforms. Recently a high-resolution structure of the catalytic domain of PDE4B has become available. In addition, the human isozymes of PDE4 (PDE4A, PDE4C, PDE4D) are highly homologous to PDE4B. Our goal is to create homology models of the various human isozymes of PDE4, which are consistent with published site, directed mutagenesis data and structure activity relationships of PDE4 inhibitors. These models will then be applied to structure-based optimization of the affinity and isozyme selectivity of beta-carboline and canthinone based alkaloids. As new structure activity relationship data is generated, the homology models will be refined in an iterative design/synthesis/screening cycle. The homology models will be used directly in structure based design by docking in candidate inhibitors into the models to predict the binding affinity and selectivity.